الفهرس 
Chapter 1: General Introduction to Optical Code Division Multiple AccessOnly 14 pages are availabe for public view
 |  | from | 293 |  |  |
| | | from | 293 | | |
Abstract
Of all kinds of multiple access techniques, optical code division multiple access (OCDMA) is considered the strongest candidate to challenge the rapid evolution of modern communication systems, due to its distinctive features required to support multimedia. However; some physical impairments such as: multiple access interference (MAI), inter symbol interference (ISI) which arises from dispersion and all forms of noise are the limiting factors that degrade the capacity and the performance of the system. This thesis has exploited the realm of digital signal processing (DSP) techniques to tackle the addressed issues and to enhance the performance of OCDMA systems. So, efficient transceivers architectures have been proposed and addressed based mainly on multiple access interference cancellation (MAIC) using; M-array of modulation schemes with employing an enhanced modified prime code (EMPC) as a spreading sequence for coding techniques. An upper bound on the bit error rate (BER) for the proposed systems is derived and compared with a lower bound on the BER from the conventional system without cancellation.
Moreover; enhanced adaptive equalizers design for wire and wireless optical communications have been suggested to combat performance degradation due to dispersion in the former and due to the adverse weather conditions due to fog and cloud in the atmospheric propagation in the latter. The proposed equalizers are based mainly on adaptive least mean square (LMS) error algorithm with adopting activity detection and tap decoupling schemes to enhance the overall performance of the equalizer. Both of the white input signals and colored input signals have been applied and the overall performance of the equalizers has been investigated. Numerical results for best parameters of equalizer design have been presented.
Furthermore; power penalty and BER have been investigated due to crosstalk (CT) which arises from the collective combination of dispersion, interference, not using
ideal (components, connectors, splices and good wave length separation). The BER performance with and without CT has been evaluated. The relation between CT versus the number of channels has been investigated.
The simulated computer results using Matlab programs have revealed that:
First, The performance of the proposed techniques of MAIC surpass the performance of the previous studies in terms of accommodating satisfactory number of simultaneous users, providing efficient power, and achieving lower BER. Such as; 1) the presented scheme of PPM with enhanced modified prime code (EMPC) achieved BER about 10-15 with affordable average number of photons per pulse = 100 (i.e. received power) as compared with respect to the conventional transceiver which has high BER equal 10-4, and needs Infinite power. 2) The addressed scheme of FSK with EMPC has provided both lower BER reached to 10-16 at repetition index j = 3 and higher bit-rate under the same conditions. Moreover, the results revealed that the FSK scheme is very power efficient and when the bit-rate is constant its network capacity can be expanded to accommodate great number of simultaneous active users with low error-rate. In addition, the proposed scheme could simplify the hardware of the receiver structure. 3) The addressed scheme of binary F-POLSK with EMPC achieved better performance than its conventional counter part of multiple tones of F-POLSK. where the results revealed that; the performance of the system using binary (F-POLSK) surpasses the performance of the system when we use multiple tones of (F-POLSK). These enhancements consist in: 1) Accommodating larger number of subscribers (capacity enhancements) at a given bit-error rates (BER) for binary (F-POLSK) than using multiple tones of (F-POLSK), where the proposed scheme supported number of users is 250 with BER = 10-9 at affordable SNR equal 16 dB as compared with respect to the lower results which have been achieved with the conventional methods at the same environments. 2) Affordable increase in the signal to noise ratio and the prime number can add more enhancements in the performance of the system without any burdens. So the proposed scheme of coding and modulation is
a promising to enhance the performance of optical CDMA system for multimedia applications.
Second, The performance of the proposed adaptive equalizers for optical fiber channel based on adaptive LMS algorithm with incorporating activity detection guidance (ADG) and tap decoupling (TD) are promising, and providing a significance improvement in the system than the performance of the conventional equalizers without adopting the ADG and TD schemes. where the enhanced scheme achieved asymptotic error performance about 10-4, the proposed scheme achieved
10-3, while the conventional scheme was 10-1. Moreover; the enhanced scheme achieved the fastest convergence rate and removed the distortion completely in addition to improving the overall performance for decreasing the noise variance to 0.001 where we have got remarkable improvement in the asymptotic error performance reached to 10-6.
Third, the performance of the presented adaptive equalizer for OWC surpasses the performance conventional equalizers. The improvement in the performance has appeared in asymptotic error performance, convergence rate, channel estimation, and the number of tap detection, where the asymptotic performance has been improved gradually for the proposed algorithm over all the other algorithms for the presented values of noise variance from 10-2 to 10-3 to 10-6 to 10-8 respectively. Thus, the simulation results for the proposed equalizer have approved the success of the proposed equalizer design to remedy the problem of ISI which arises from the impairments of wireless channel.
Fourth, The simulation of power penalty and BER investigation due to crosstalk revealed that; the power penalty due to the effect of thermal noise in PIN photodiode is more harmful than the effect of spontaneous beat noise in avalanche photodiode, and they approved that; the intraband CT did not affect the received signal as long as it is less than -25 dB. For ideal system design the CT will equal to zero and hence we can get lower BER with lower input power than the case when the crosstalk exists. BER
versus input power with and without CT has been carried out in the simulation results, and the power penalty for each crosstalk has been calculated. Finally, the relation between CT against the number of channels with different hops has been investigated, and the results revealed that; increasing the number of channels and the number of hops will increase the CT which in turn increases the power penalty and BER.